Transcript ppt - 503 KB

Time/frequency analysis with COROT:
what do we want to do?
F. Baudin & E. Michel
Time/frequency analysis : any method providing an information
on the temporal behaviour of the star’s vibrations.
For example : wavelet transform, Gabor transform, curvelet
analysis, matching pursuit…
Baudin et al. A&A, 1996
Temporal behaviour of modes = information on the
mechanism of excitation of the modes
Several cases:
• solar like oscillators : excitation by convection
• « classical » pulsators (d Scuti, b Cepheids…) : excitation
by k mechanism
Several questions:
• are solar like oscillators really solar like?
• what is the saturation mechanism in classical pulsators?
• other mechanism of excitation? (roAp, SPB, g Dor…)
• etc…
Solar like oscillators
Do they behave as expected? (the Sun’s case is more or less
well understood but it’s only one case)
=> exploratory program based on the long runs of COROT
(150 days of continuous observation is something unique at
present time, possibility of statistic studies on power variations
and comparison of the different COROT targets)
=> information on the convection, which can behave in
different manners in different stars (thickness of the CZ,
efficiency of convection)
Classical pulsators
k mechanism increases the amplitude of the modes until
« saturation »
- What is this mechanism of saturation ?
- Is it the same in all classical pulsators ?
- Is there evidence of pumping (coupling) of the mode’s power
by another mode?
- Is there any modulation by the rotation or magnetic activity
(or something else) ?
- etc…
=> Clues for the determination of this unknown saturation
mechanism based on the continuous observations of COROT
Other pulsators
roAp, SPB, sdB, g Dor:
Unknown or not well understood mechanism of excitation
=> Clues for the determination of this unknown/not well
understood mechanism based on the continuous observations
of COROT
Solar like oscillators
Is there some coupling between modes in solar like oscillators?
Modes are independent (orthogonal) in linear approximation,
but do we see evidences of coupling and thus the necessity for
non linear approach even in the solar like case ?
Baudin et al. A&A, 1994
All pulsators
Existence of another excitation mechanism? (flare related)
Solar case:
Ici figure Soho14
Baudin & Finidori, Soho14, 2004
Baudin et al. in prep.
All pulsators
Existence of another excitation
mechanism? (flare related)
GD358 (white dwarf) case:
Kepler et al. A&A 2003
How to do that?
Wavelet analysis, curvelet analysis, matching pursuit…
All have to deal with Heisenberg relation in signal analysis:
DT Dn > cte
Solar case:
l = 0/l = 2 Dn ~ 10mHz
=> DT ~ 1 day
How to do that?
DT ~ 1 day?
Presence of noise => spurious power temporal variations
DT SNR > cte
Need for an estimation of the
possible resolution in time
and frequency and trade off
with noise:
Use of simulated data +
expected mode spectrum
structure
How to do that?
(in practice)
Let’s coordinate ourselves!
• let’s talk
• frederic.baudin @ ias.u-psud.fr / eric.michel @ obspm.fr